應用多目標遺傳演算法於網際網路廣告採買策略最佳化

Applying Multi-Objective Genetic Algorithms for Optimal Internet Advertisement Purchase

陳有賢(Yu-Hsien Chen)

指導教授 ： 邱昭彰

元智大學/資訊學院/資訊管理學系/碩士(2003年)

摘要

隨著網際網路的普及，企業或廣告主紛紛藉由在網路上刊登廣告來獲取更大的收入或是利潤。然而，大部分的廣告主都是經由廣告代理商來進行網際網路廣告的採買。但是這普遍存在幾個問題，第一：由於網路調查數據各家不一，以致於常常造成市場上的混淆，因此廣告代理商所擬訂出來的組合是真正對廣告主有利？第二：廣告代理商甚至是媒體代表在擬定媒體組合策略時，通常只以點閱率或是曝光次數等等類似的單一目標來進行，而所擬訂出來的策略真的是符合廣告主所需要的？因此，本研究將針對網際網路廣告採買最佳化，提出一解決方法。廣告採買最佳化是指在眾多網路媒體中應該要在哪些廣告版面購買多少量的廣告，使得企業或是廣告主的行銷目標可以獲得最大的效益。首先我們會建立一個網際網路廣告採買最佳化的模型並且以某網路收視調查公司之資料為基礎，整合在模型中。為了解決這樣的最佳化問題，我們使用多目標最佳化的技術來求解網際網路廣告採買最佳化的模型。其中多目標最佳化的技術為多目標遺傳演算法。

關鍵字

網路廣告 ； 廣告採買最佳化 ； 多目標最佳化 ； 遺傳演算法 ； 多目標遺傳演算法

並列摘要

It is able to directly interaction with customer and the fee lower then traditional media when the companies want to make more benefit by advertising in the Internet. However, most advertisers who purchase media of cyberspace via advertise agency. There are some problems that one is the media planner of advertising agency proceed media mix plan just use the click rate of web site or use fix index about operator web site. Second is the survey data always not coincidence that published by each web site. So our research attempts to solve the problems of advertisement purchase optimization. Advertisement purchase optimization means finding a portfolio of advertisement purchase that the enterprise will gain the greatest benefit. First, we establish the advertisement purchase optimization model based on the survey data of website that provide by one specific marketing survey company and information form advertisement analysis of advertiser. In order to solve such optimization problem with multiple objectives, we apply multiple-objective optimization to solve the advertisement purchase optimization model. The optimization technique of multiple-objective optimization we apply is multi-objective genetic algorithm.

並列關鍵字

Internet Advertising ； Advertisement purchase Optimization ； Multiple objective optimization ； Genetic Algorithms ； Multi-objective Genetic Algorithms

參考文獻

[1] 丁雅方, “多屬性效用模式與基因演算法在網際網路媒體行銷組合最佳化之決策輔助” ,元智大學,碩士論文,1998.

[10] Ahuja, R. K., Orlin, J. B. and Tiwari, A., “A Greedy Genetic Algorithm for The Quadratic Assignment Problem,” Computers and Operations Research, Volume: 27, Issue: 10, pp. 917-934, September 2000,

[11] Apte, G., Weiss, S., “Data Mining with Decision Trees and Decision Rules,” Future Generation Computer Systems, Volume: 13, pp. 197-210, 1997.

[12] Bo, Z., Ho, S. L., Lin, G. and Masatoshi, T. ,“Robust Turning of Fix-Structure for Disk Drives Using Statistical Model and Multi-Objective Genetic Algorithms,” in Proceedings of American Control Conference, pp 2773-2778, 2001.

[13] Botes, J. H. F., Bosch, D. J. and Oosthuizen, L. K., “A Simulation and Optimization Approach for Evaluating Irrigation Information,” Agricultural Systems, Vol. 51, Issue. 2, pp. 165-183, June 1996.